INDEX

Explanations

words related to qualities like smoothness, plausibility, unpredictability, and complexity

oai_token-act-pair · gpt-3.5-turbo

abstract nouns that convey qualities or states of being

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

nard

-0.85

ergy

-0.72

inav

-0.63

rgb

-0.62

iverse

-0.61

inis

-0.59

som

-0.59

 Crest

-0.56

members

-0.56

JECT

-0.55

POSITIVE LOGITS

iness

1.15

ously

0.92

lessness

0.89

icity

0.83

worthiness

0.82

ality

0.78

acies

0.78

fulness

0.78

 afforded

0.75

uous

0.74

Activations Density 0.077%

words related to qualities like smoothness, plausibility, unpredictability, and complexity

abstract nouns that convey qualities or states of being

No Comments

No Known Activations